Alpha-Beta Pruning Example

Alpha-Beta Pruning Example
Alpha-Beta Pruning Idea

Some of the branches of the game tree won't be taken if playing against a smart opponent. Use pruning to ignore those branches. While doing DFS of game tree, keep track of:
alpha at maximizing levels (computers move)
highest SBE value seen so far (initialize to -infinity) is lower bound on state's evaluation
beta at minimizing levels (opponents move)

lowest SBE value seen so far (initialize to +infinity) is higher bound on state's evaluation
Alpha-Beta Pruning Idea

Beta cutoff pruning occurs when maximizing if childs alpha >= parent's beta Why stop expanding children?
opponent won't allow computer to take this move
Alpha cutoff pruning occurs when minimizing if parent's alpha >= childs beta Why stop expanding children?
computer has a better move than this
Alpha-Beta Search Example

minimax(A,0,4) alpha initialized to -infinity
Expand A? Yes since there are successors, no cutoff test for root A A =C G H
3
max B F N
4
Call Stack D
0
E K L
2
-5
J Q R
0
M U V
O W
-3
P
9
-6
S
3
T
5
-7
-9
-5

minimax(B,1,4) beta initialized to +infinity
Expand B? Yes since As alpha >= Bs beta is false, no alpha cutoff A Call Stack D
0
max B B =+ min F N
4 -5
=-
C H
3
E K L
2
J Q R
0
M U V
O W
-3
P
9
-6
S
3
T
5
-7
-9
-5
B A

minimax(F,2,4) alpha initialized to -infinity
Expand F? Yes since Fs alpha >= Bs beta is false, no beta cutoff A =- C H

3
max
=+
Call Stack D
0
E K L
2
min max
4
F F =- N W
-3
-5
J Q R
0
M U V
O X
P
9
-6
S
3
T
5
-7
-9
F B A
-5

minimax(N,3,4) evaluate and return SBE value
max
=+
A =- C D
0
Call Stack E K R
0
min max
4
F =- N W
-3
-5
H
3
J Q S
3
L
2
M U V N F B A
O X
P
9
-6
T
5
-7
-9
-5
green: terminal state

back to minimax(F,2,4) alpha = 4, since 4 >= -infinity (maximizing)
Keep expanding F? Yes since Fs alpha >= Bs beta is false, no beta cutoff A =- C H
3
max
=+
Call Stack D
0
E K L
2
min max N
4
=4 =-
-5
J Q R
0
M U V
O W
-3
P
9
-6
S
3
T
5
-7
-9
-5
F B A

minimax(O,3,4) beta initialized to +infinity
Expand O? Yes since Fs alpha >= Os beta is false, no alpha cutoff A =- C H

3
max
=+
Call Stack D
0
E K L
2
min max N
4
=4
-5
J Q R
0
M U V O F B A
O O =+ W
-3 -5
P
9
-6
S
3
T
5
-7
-9
min

minimax(W,4,4) evaluate and return SBE value
max
=+
A =- C D
0
Call Stack E K R
0
min max N
4
=4
-5
H
3
J Q S
3
L
2
M U V
=+
P
9
-6
T
5
-7
-9
min
W
-3
-5
blue: non-terminal state (depth limit)
W O F B A

back to minimax(O,3,4) beta = -3, since -3 <= +infinity (minimizing)
Keep expanding O? No since Fs alpha >= Os beta is true: alpha cutoff A =- C H

3
max
=+
Call Stack D
0
E K L
2
min max N
4
=4
-5
J Q R
0
M U V O F B A
=-3 =+
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5

Why? Smart opponent will choose W or worse, thus O's upper bound is 3. Computer already has better move at N.
max
=+
A =- C D
0
Call Stack E K R
0
min max N
4
=4
-5
H
3
J Q S
3
L
2
M U V O F B A
=-3
P
9
-6
T
5
-7
-9
min
W
-3
-5
red: pruned state

back to minimax(F,2,4) alpha doesnt change, since -3 < 4 (maximizing)
Keep expanding F? No since no more successors for F A =- C H

3
max
=+
Call Stack D
0
E K L
2
min max N
4
=4
-5
J Q R
0
M U V
=-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
F B A

back to minimax(B,1,4) beta = 4, since 4 <= +infinity (minimizing)
Keep expanding B? Yes since As alpha >= Bs beta is false, no alpha cutoff A =- C G H
3
max
=+ =4
Call Stack D
0
E K L
2
min
=4
-5
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
B A

minimax(G,2,4) evaluate and return SBE value
max
=4
A =- C D
0
Call Stack E K R
0
min
=4
-5
H
3
J Q S
3
L
2
M U V
max N
4 =-3
P
9
-6
T
5
-7
-9
min
W
-3
-5
G B A

back to minimax(B,1,4) beta = -5, since -5 <= 4 (minimizing)
Keep expanding B? No since no more successors for B A =- C G H

3
max
=-5 =4
Call Stack D
0
E K L
2
min
=4
-5
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
B A

back to minimax(A,0,4) alpha = -5, since -5 >= -infinity (maximizing)
Keep expanding A? Yes since there are more successors, no cutoff test A Call Stack D
0
max
=-5
=-5 =
C G H
3
E K L
2
min
=4
-5
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5

minimax(C,1,4) beta initialized to +infinity
Expand C? Yes since As alpha >= Cs beta is false, no alpha cutoff A Call Stack D
0
max
=-5
=-5 =
min
=4
C C =+
-5
E K L
2
H
3
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
C A

minimax(H,2,4) evaluate and return SBE value
max
=-5
=-5 = =+
Call Stack D
0
C I
E K L
2
min
=4
-5
H
3
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
H C A

back to minimax(C,1,4) beta = 3, since 3 <= +infinity (minimizing)
Keep expanding C? Yes since As alpha >= Cs beta is false, no alpha cutoff A Call Stack D
0
max
=-5
=-5 = =+ =3
C I
E K L
2
min
=4
-5
H
3
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
C A

minimax(I,2,4) evaluate and return SBE value
max
=-5
=-5 = =3
Call Stack D
0
C I
E K L
2
min
=4
-5
H
3
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
I C A

back to minimax(C,1,4) beta doesnt change, since 8 > 3 (minimizing)
Keep expanding C? Yes since As alpha >= Cs beta is false, no alpha cutoff A Call Stack D
0
max
=-5
=-5 = =3
C I
E K L
2
min
=4
-5
H
3
J Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
C A

minimax(J,2,4) alpha initialized to -infinity
Expand J? Yes since Js alpha >= Cs beta is false, no beta cutoff A Call Stack D
0
max
=-5
=-5 = =3
C I
E K L
2
min
=4
-5
H
3
J J =Q R
0
M U V
max N
4 =-3
P
9
-6
S
3
T
5
-7
-9
min
W
-3
-5
J C A

minimax(P,3,4) evaluate and return SBE value
max
=-5
=-5 = =3
Call Stack D
0
C I
E K L
2
min
=4
-5
H
3
max N
4 =-3
=-
M U V P J C A
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5

back to minimax(J,2,4) alpha = 9, since 9 >= -infinity (maximizing)
Keep expanding J? No since Js alpha >= Cs beta is true: beta cutoff A Call Stack D
0
max
=-5
=-5 = =3
C I
E K L
2
min
=4
-5
H
3
=9 =-
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5
red: pruned states
J C A

Why? Computer will choose P or better, thus J's lower bound is 9. Smart opponent wont let computer take move to J (since opponent already has better move at H). A Call Stack D
0
max
=-5
=-5 = =3
C I
E K L
2
min
=4
-5
H
3
=9
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5
red: pruned states
J C A

back to minimax(C,1,4) beta doesnt change, since 9 > 3 (minimizing)
Keep expanding C? No since no more successors for C A Call Stack D

0
max
=-5
=-5 = =3
C I
E K L
2
min
=4
-5
H
3
=9
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5
C A

back to minimax(A,0,4) alpha = 3, since 3 >= -5 (maximizing)
0
max
=-5
=-5 =3 =3
C I
E K L
2
min
=4
-5
H
3
=9
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5

minimax(D,1,4) evaluate and return SBE value
max
=-5
=3 = =3
Call Stack D
0
C I
E K L
2
min
=4
-5
H
3
=9
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5
D A

back to minimax(A,0,4) alpha doesnt change, since 0 < 3 (maximizing)
0
max
=-5
=3 = =3
C I
E K L
2
min
=4
-5
H
3
=9
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5

How does the algorithm finish searching the tree?
max
=-5
=3 = =3
Call Stack D
0
C I
E K L
2
min
=4
-5
H
3
=9
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5

Stop Expanding E since A's alpha >= E's beta is true: alpha cutoff Why? Smart opponent will choose L or worse, thus E's upper bound is 2. Computer already has better move at C. max
=-5 =3 = =3
Call Stack D
0 =2
C I
E L
2
min
=4
-5
H
3
=9
=5
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5

Result: Computer chooses move to C.
max
=-5
=3 = =3
Call Stack D
0 =2
C I
E L
2
min
=4
-5
H
3
=9
=5
M U V
max N
4 =-3
P
9
-6
R
0
S
3
T
5
-7
-9
min
W
-3
-5
green: terminal states, red: pruned states blue: non-terminal state (depth limit)

Alpha-Beta Pruning Example

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Alpha-Beta Pruning Example

Hochgeladen von

Copyright:

Verfügbare Formate

Alpha-Beta Pruning Example

Alpha-Beta Pruning Idea

beta at minimizing levels (opponents move)

Alpha-Beta Pruning Idea

Alpha-Beta Search Example

Alpha-Beta Search Example

Alpha-Beta Search Example

Expand F? Yes since Fs alpha >= Bs beta is false, no beta cutoff A =- C H

Alpha-Beta Search Example

green: terminal state

Alpha-Beta Search Example

Alpha-Beta Search Example

Expand O? Yes since Fs alpha >= Os beta is false, no alpha cutoff A =- C H

Alpha-Beta Search Example

blue: non-terminal state (depth limit)

Alpha-Beta Search Example

Keep expanding O? No since Fs alpha >= Os beta is true: alpha cutoff A =- C H

Alpha-Beta Search Example

red: pruned state

Alpha-Beta Search Example

Keep expanding F? No since no more successors for F A =- C H

Alpha-Beta Search Example

Alpha-Beta Search Example

green: terminal state

Alpha-Beta Search Example

Keep expanding B? No since no more successors for B A =- C G H

Alpha-Beta Search Example

Alpha-Beta Search Example

Alpha-Beta Search Example

green: terminal state

Alpha-Beta Search Example

Alpha-Beta Search Example

green: terminal state

Alpha-Beta Search Example

Alpha-Beta Search Example

Alpha-Beta Search Example

green: terminal state

Alpha-Beta Search Example

red: pruned states

Alpha-Beta Search Example

red: pruned states

Alpha-Beta Search Example

Keep expanding C? No since no more successors for C A Call Stack D

Alpha-Beta Search Example

Alpha-Beta Search Example

green: terminal state

Alpha-Beta Search Example

Alpha-Beta Search Example

Alpha-Beta Search Example

Alpha-Beta Search Example

Das könnte Ihnen auch gefallen